18F-Fluorodeoxysorbitol PET for noninvasive detection of invasive mold infections: preclinical and first-in-human studies

Ruiz-Gonzalez, Carlos E.; Nino-Meza, Oscar J.; Singh, Medha; Masias-Leon, Yuderleys; Kronenberg, Amy; Shambles, Madelynn; Chen, Xueyi; Tucker, Elizabeth W.; Lodge, Martin A.; Carroll, Laurence S.; Cooke, Kenneth R.; Kates, Olivia S.; Shoham, Shmuel; Zhang, Sean X.; Jain, Sanjay K.

doi:10.1038/s41467-025-61700-6

Download PDF

Article
Open access
Published: 10 July 2025

¹⁸F-Fluorodeoxysorbitol PET for noninvasive detection of invasive mold infections: preclinical and first-in-human studies

Carlos E. Ruiz-Gonzalez^1,2^na1,
Oscar J. Nino-Meza^1,2^na1,
Medha Singh^1,2^na1,
Yuderleys Masias-Leon ORCID: orcid.org/0000-0002-8333-7180^1,2,
Amy Kronenberg ORCID: orcid.org/0009-0000-4318-0628¹,
Madelynn Shambles¹,
Xueyi Chen ORCID: orcid.org/0000-0003-1574-7759^1,2,
Elizabeth W. Tucker ORCID: orcid.org/0000-0002-0971-3635^1,3,
Martin A. Lodge⁴,
Laurence S. Carroll⁴,
Kenneth R. Cooke⁵,
Olivia S. Kates⁶,
Shmuel Shoham⁶,
Sean X. Zhang⁷ &
…
Sanjay K. Jain ORCID: orcid.org/0000-0001-9620-7070^1,2,4

Nature Communications volume 16, Article number: 6395 (2025) Cite this article

4979 Accesses
2 Citations
8 Altmetric
Metrics details

Subjects

Abstract

Invasive mold infections are a major cause of mortality in immunosuppressed and cancer patients. Diagnosis is challenging, requiring invasive procedures or reliance on fungal biomarkers with limited sensitivity and an inability to detect non-Aspergillus molds. Here, we perform whole-body ¹⁸F-fluorodeoxysorbitol (¹⁸F-FDS) positron emission tomography (PET) in nine prospectively enrolled patients with high-suspicion of invasive mold infections (eventually confirmed using culture or molecular assays, n = 4) or other pathologies (n = 5) with localization of ¹⁸F-FDS PET signal to infection sites as the primary outcome (NCT05611892). ¹⁸F-FDS PET (120 or 180 min after injection), rapidly detects and localizes invasive pulmonary and cerebral infections due to Aspergillus, non-Aspergillus (galactomannan-negative), or azole-resistant molds, and differentiates them from sterile inflammation or cancer. Moreover, ¹⁸F-FDS selectively and rapidly accumulates intracellularly in a range of clinically relevant molds, including azole-resistant molds, via a saturable process. In animals, ¹⁸F-FDS PET is able to detect and localize pulmonary and cerebral aspergillosis, as well as rhinosinusal infections due to Aspergillus, Rhizopus, and Mucor, confirming the clinical data. ¹⁸F-FDS can be easily synthesized from ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG), which is widely available, and represents a promising, noninvasive diagnostic tool for detecting, localizing and monitoring of invasive mold infections throughout the body.

In vivo imaging of invasive aspergillosis with ¹⁸F-fluorodeoxysorbitol positron emission tomography

Article Open access 08 April 2022

Retrospective study on clinical value and optimal use of [¹⁸F] FDG PET/CT for inflammation of unknown origin in Japanese patients

Article Open access 15 November 2024

Kit-based synthesis of 2-deoxy-2-[¹⁸F]-fluoro-d-sorbitol for bacterial imaging

Article 22 October 2021

Introduction

Invasive mold infections are an increasing global public health concern, affecting multiple organ systems (e.g., lung, brain), and associated with mortality rates as high as 85%^1,2,3. The rising incidence of invasive mold infections is attributed to the expanding population of immunosuppressed patients³. The COVID-19 pandemic has also led to a dramatic increase in invasive mold infections, particularly pulmonary aspergillosis and mucormycosis^4,5. Aspergillus spp. remains the major cause of invasive mold infections in hematologic cancer patients and transplant recipients. However, other opportunistic molds, which are even more challenging to diagnose and treat⁶, are on the rise due to several reasons, including wide-spread Aspergillus-targeted antifungal prophylaxis^7,8. In 2022, the World Health Organization (WHO) categorized Aspergillus and Mucorales molds as Critical and High Priority Group pathogens, respectively⁹. Prompt identification of invasive mold infections and initiation of appropriate treatment is critical for improving patient outcomes.

When infections occur in relatively inaccessible sites, minimally invasive clinical samples (blood, urine, stool, or cerebrospinal fluid) can often yield non-diagnostic results or are otherwise insensitive for deep-seated infections. Although major strides have been made in the development of microbial cell-free DNA detection for infections^10,11, clinically available biomarkers, namely galactomannan and 1,3-beta-D-glucan, have limited sensitivity and specificity for Aspergillus spp., and galactomannan is unable to reliably detect non-Aspergillus molds. Specifically, the overall sensitivity of serum / plasma galactomannan using the most current (2020) European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium (EORTC / MSGERC) guidelines is 65-71%, but lower in those with hematologic malignancies (58%), hematopoietic stem cell transplantation (65%), and non-neutropenic patients (24%)^12,13. While galactomannan from the bronchoalveolar lavage (BAL) has a higher sensitivity (75–86%), BAL is an invasive process and cannot be performed in many cancer patients at risk for bleeding complications.

Importantly, all clinically available imaging tools such as radiography, ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI) rely on structural changes in anatomy or tissue morphology that are often delayed relative to the disease process, are non-specific, and reflect a combination of the infection and the host inflammatory response^14,15,16. CT findings of pulmonary aspergillosis include the halo sign (nodule with surrounding ground-glass opacity), consolidations, and cavitary lesions, but these findings may also be present with other pathological processes¹⁷. Although ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) is highly sensitive for detecting infectious foci, including Aspergillosis^18,19,20, it also reflects a combination of the infection and inflammation. Therefore, currently available imaging tools rely on host responses to infection that could be significantly altered in immunocompromised host and none can reliably differentiate infectious sites from sterile inflammatory or cancerous lesions^14,15,16. Moreover, ¹⁸F-FDG PET may also have limited accuracy for detecting cerebral fungal infections¹⁹, presumably due to the high background signal in the brain. Additionally, biomarkers cannot provide spatial information on the location or extent of infection. Therefore, definitive diagnosis of invasive mold infections almost always requires invasive procedures such as BAL or deep tissue biopsy to obtain a relevant clinical sample, which is challenging, dangerous (e.g., brain biopsy), and delays the diagnosis. Because of these challenges, empiric antifungal treatment (prior to establishing a confirmatory diagnosis) is the norm in patients with suspected invasive mold infections. However, treatment can decrease the sensitivity of culture-based tests performed later. The development of rapid, whole-body, noninvasive, and specific diagnostics, able to detect invasive mold infections in patients who are often immunosuppressed and receiving empiric treatment, would be a crucial advancement in the field and could become the clinical standard of care. Although, there have been several efforts to develop pathogen-specific imaging approaches for Aspergillus using siderophores, labeled antibodies and sugars^{16,21,22,23,24,25,26,27,28,29,30}, none have been prospectively studied in patients (other than a case report³¹), or applied to other mold infections. Moreover, there are some challenges in the use of labeled antibodies, namely slower clearance, potential for reduced penetration in necrotic tissues, as well as more complicated regulatory pathways (versus small molecules) for clinical translation. Therefore, there is an urgent need for sensitive, whole-body, noninvasive tools to diagnose invasive mold infections.

We have previously described ¹⁸F-fluorodeoxysorbitol (¹⁸F-FDS), a radioanalog of sorbitol (sugar alcohol), as a PET tracer to selectively image infections due to the Enterobacterales group of bacteria (e.g., Escherichia coli)^32,33. However, Aspergillus can also utilize sorbitol as a carbon source^34,35, and a recent study has shown that ¹⁸F-FDS PET can detect invasive pulmonary aspergillosis in animal models²⁷. Our overall goals were to evaluate ¹⁸F-FDS PET to specifically detect and localize invasive mold infections by a wide range of clinically relevant mold-fungi in animal models and first-in-human studies.

Here, we perform whole-body ¹⁸F-FDS PET/CT in nine prospectively enrolled patients with either invasive mold infections (n = 4 patients) or other pathologies (n = 5 patients) (NCT05611892). We demonstrate that ¹⁸F-FDS PET can rapidly detect and localize invasive pulmonary and cerebral mold infections due to drug-susceptible or azole-resistant molds, and differentiate them from sterile inflammation and cancer. We also show that ¹⁸F-FDS selectively accumulates in a range of clinically relevant molds, but not by heat killed fungi or mammalian cells. Importantly, ¹⁸F-FDS is rapidly metabolized intracellularly by molds, via a saturable process. Tissue histology from eight unique patients with pulmonary, cerebral or rhinosinusal invasive mold infections, demonstrates the lesional biomass occupied by molds is ~50-fold higher than for bacterial infections. In animal studies, ¹⁸F-FDS PET/CT is able to detect and localize pulmonary and cerebral aspergillosis, as well as rhinosinusal infections due to several molds, confirming the clinical data. Importantly, ¹⁸F-FDS can be easily synthesized from commercially available ¹⁸F-FDG³⁶, the most widely utilized PET tracer globally, by a simple one-step kit-based method, without the need for specialized radiosynthesis and purification facilities³⁷, allowing on-demand synthesis and global availability of ¹⁸F-FDS.

Results

¹⁸F-FDS PET/CT selectively detects invasive mold infections

We prospectively enrolled nine newly identified patients with either invasive mold infections (n = 4) using the EORTC/MSGERC consensus definitions (Table S1) or uninfected controls with other pathologies (n = 5) (NCT05611892). Confirmatory diagnosis of mold infections was established in all four patients, using culture or molecular assays (Fig. 1, Table S2). Sterile inflammatory pathologies (n = 4) were determined clinically and/or by tissue biopsies and oncologic diagnosis (n = 1; breast cancer) was made via tissue biopsy (Fig. 1, Table S2)³³. The median age was 59 years (range, 37 to 71 years), and 44% (4 of 9) were female. Importantly, the enrolled patients were representative of patients hospitalized in a tertiary care setting with several co-morbidities (Table S2). Whole-body ¹⁸F-FDS PET/CT performed in accordance with U.S. FDA guidelines was well tolerated in all patients. ¹⁸F-FDS kinetics were consistent with prior studies³³, demonstrating rapid renal / hepatobiliary elimination and low background in unaffected brain and lung tissues.

To quantify the PET signal, spherical volumes of interest (VOIs) were drawn at the sites of pathology and unaffected sites of the same tissue to calculate the target-to-nontarget ratio (TNT). ¹⁸F-FDS PET signal was significantly higher at the sites of invasive mold infections with a TNT of 5.79 [interquartile range (IQR), 3.18 to 15.20] versus sterile inflammatory pathologies or cancer with a TNT of 1.40 (IQR, 1.01 to 1.93) (P = 0.008) (Fig. 2). Importantly, while the ¹⁸F-FDG PET clearly demonstrated the breast cancer lesion, no signal was noted on ¹⁸F-FDS PET (Fig. S1). Using a previously defined TNT cutoff of 3.0³³, ¹⁸F-FDS PET was able to detect and localize invasive pulmonary and cerebral infections in all patients with mold infections [Aspergillus, non-Aspergillus (galactomannan-negative) molds, or azole-resistant Aspergillus calidoustus)], despite having received up to seven days of antifungal treatments by the time of imaging. Moreover, lesions in all control patients had a TNT ratio of <3.

Fig. 2: Quantification of 18F-FDS PET signal in patients with mold infections. — **Fig. 2: Quantification of ¹⁸F-FDS PET signal in patients with mold infections.**

For pulmonary infections, ¹⁸F-FDS PET signal was significantly higher at infection sites with a TNT of 3.40 (IQR, 3.10 to 14.05) compared to sites with sterile inflammatory pathologies with a TNT of 1.79 (IQR, 1.57 to 2.17) (P < 0.001) (Fig. 3, S2–3, S6). Similarly, the ¹⁸F-FDS PET signal was significantly higher at the sites of invasive cerebral mold infections with a TNT of 6.39 (IQR, 5.19 to 16.92) (P = 0.007) (Fig. 4, S4–6). Importantly, ¹⁸F-FDS PET was able to identify a cerebral lesion in patient 4 with Cladophialophora bantiana pulmonary infection, which was previously missed on a brain MRI performed for clinical reasons (Fig. S4).

Fig. 3: Thoracic 18F-FDS PET/CT imaging in patients. — **Fig. 3: Thoracic ¹⁸F-FDS PET/CT imaging in patients.**

Fig. 4: Cranial 18F-FDS PET imaging in patients. — **Fig. 4: Cranial ¹⁸F-FDS PET imaging in patients.**

¹⁸F-FDS uptake by clinically relevant molds

In vitro studies were performed to determine ¹⁸F-FDS uptake by a range of clinically relevant molds (Fig. 5a). Rapid ¹⁸F-FDS accumulation was noted in the live strains, whereas no uptake was noted in the heat killed controls (Fig. 5b–d) or in mammalian cells (THP-1 macrophages) (Fig. S7). Competitive inhibition tests conducted by co-incubation with increasing concentrations of unlabeled sorbitol decreased the ¹⁸F-FDS uptake (Fig. 5g, e), suggesting a saturable mechanism. Finally, all the 30 randomly selected clinical isolates, including Aspergillus spp., Rhizopus spp., Mucor spp., and other clinically relevant molds (Lichtheimia corymbifera, Syncephalastrum racemosum, Cunninghamella bertholletiae, Cladophialophora bantiana, and Fusarium solani), along with azole-resistant molds demonstrated robust ¹⁸F-FDS uptake. These findings confirm that sorbitol uptake is conserved across a range of clinically relevant molds (Fig. 5h).

Fig. 5: 18F-FDS uptake by clinically relevant molds. — **Fig. 5: ¹⁸F-FDS uptake by clinically relevant molds.**

Finally, we performed studies to elucidate the metabolism of ¹⁸F-FDS by molds. We co-incubated actively growing Aspergillus fumigatus with ¹⁸F-FDS and analyzed the cytoplasmic fraction from the lysed fungal pellet by radio-thin layer chromatography (radio-TLC). ¹⁸F-FDS was rapidly metabolized intracellularly by Aspergillus in a manner similar to the intracellular metabolism of ¹⁸F-FDS in E. coli (Fig. S8).

Mold biomass occupies a significant proportion of the infected lesion

The in vitro uptake of ¹⁸F-FDS is lower in molds compared to E. coli (Fig. S7). However, in vivo visualization of infections by ¹⁸F-FDS PET is dependent on the total tracer uptake within the infected lesion. We therefore compared the biomass within infectious lesions occupied by molds and bacteria (Fig. 6). Infected tissue samples from eight unique patients with pulmonary, cerebral or rhinosinusal invasive Aspergillus infections from the Johns Hopkins Hospitals (patient characteristics are outlined in Table S3), or publicly available digitized slides of bacterial infections in patients³⁸, were analyzed. Tissue from animals infected with Aspergillus, Rhizopus, Mucor or E. coli were also analyzed. High resolution images are shown in Fig. S9. The percentage of the infected lesion volume occupied by the mold or bacteria was calculated using automated segmentation (Fig. 6, S10). Data from human tissues demonstrated that the biomass occupied by Aspergillus was 11.71% (IQR 7.85 to 13.03%), which is ~50-fold higher than for bacterial infections at 0.24% (IQR 0.10 to 0.38%) (P < 0.001). Similar findings were observed in tissues obtained from animal studies (P = 0.016).

**Fig. 6: Tissue histology from human and animal studies.**

¹⁸F-FDS PET localizes fungal infections in animal models

Next, we developed several clinically relevant mouse models of Aspergillus, Rhizopus, and Mucor infections and evaluated the ability of ¹⁸F-FDS PET to detect and localize the mold infections. Immunosuppressed mice were used for all studies. For pulmonary aspergillosis, mice were infected intratracheally with increasing doses of A. fumigatus conidia and the pulmonary ¹⁸F-FDS PET signal was measured, demonstrating a PET signal proportional to the infection burden (Fig. S11). Thereafter, we evaluated ¹⁸F-FDS PET in uninfected animals, animals with pulmonary aspergillosis and those with lipopolysaccharide (LPS)-induced sterile pulmonary inflammation (Fig. S12). Although substantial pulmonary disease was noted on CT and ¹⁸F-FDG PET in mice with sterile inflammation, no ¹⁸F-FDS PET signal was noted in these mice (Fig. S12b). However, mice with pulmonary aspergillosis demonstrated ¹⁸F-FDS PET signal, which colocalized with the lesions noted on CT (Fig. S12c). The pulmonary ¹⁸F-FDS PET signal was substantially higher in the mice with pulmonary aspergillosis versus those with sterile inflammation (Fig. S12g, P = 0.003). By contrast, ¹⁸F-FDG PET, which measures inflammation and is not specific for infection, could not distinguish mice with pulmonary aspergillosis versus those with sterile inflammation (Fig. S12h, P = 0.338). Postmortem pulmonary tissue histology demonstrated Aspergillus hyphae in the mice with pulmonary aspergillosis (Fig. S12i). We also performed studies with Pseudomonas, that can cause pneumonia in immunosuppressed patients. The pulmonary ¹⁸F-FDS PET signal was substantially higher in the mice with pulmonary aspergillosis versus those with Pseudomonas pneumonia (Fig. S13, P = 0.002). Importantly, the ¹⁸F-FDS PET signal from P. aeruginosa pneumonic lesions were no different than those noted with sterile pulmonary inflammation due to LPS. Consistent with our prior studies³², in vitro uptake of ¹⁸F-FDS by Pseudomonas was substantially lower than the uptake by molds (P = 0.003) (Fig. S13c).

We also developed mouse models of rhinosinusal mold infections due to A. fumigatus, Rhizopus arrhizus and Mucor circinelloides, and performed ¹⁸F-FDS PET, which was able to detect and localize the infection sites with significantly higher PET signals noted in mice with mold infections versus those with LPS-related systemic inflammation (Fig. S14, P < 0.002). Postmortem ex vivo biodistribution confirmed the imaging findings and tissue histology demonstrated fungal hyphae in the infected mice (Fig. S14e-f). Further, we performed ¹⁸F-FDS PET/CT in Aspergillus-infected mice with rhinosinusal infection to assess response to antifungal treatments. ¹⁸F-FDS PET signal in voriconazole treated (for 1 week) mice was significantly lower compared to untreated mice (P = 0.011) (Fig. S15).

Finally, we developed a mouse model of invasive cerebral aspergillosis, utilizing direct intracerebral infection with A. fumigatus. ¹⁸F-FDS PET was able to detect and localize the infection sites with significantly higher PET signal in mice with invasive cerebral aspergillosis compared to mice with LPS-related systemic inflammation (Fig. 7a–d, P < 0.001). Of note, the “ring enhancing” PET uptake pattern in mice (Fig. 7c) was similar to that noted in the patient with invasive cerebral aspergillosis (Fig. 4a), which is well described clinically^39,40, and likely represents a lack of blood supply and dead tissues in the necrotic core of the lesion. Postmortem ex vivo biodistribution confirmed the imaging findings and tissue histology demonstrated fungal hyphae in the infected mice (Fig. 7e, f). Additionally, the ¹⁸F-FDS PET signal colocalized with the lesions noted on the MRI (Fig. 7h).

**Fig. 7: Mouse model of invasive cerebral aspergillosis.**

Discussion

Here, we describe results from a prospective study evaluating ¹⁸F-FDS PET/CT to specifically detect and localize invasive mold infections (and differentiate them from other pathologies) in a patient population representative of a tertiary care hospital setting within seven days of initiation of antifungal treatments. Although larger studies are needed to better assess the diagnostic accuracy, by applying a previously defined³³ TNT cutoff of 3, in this small study, ¹⁸F-FDS PET correctly classified all patients with and without invasive mold infections.

Differentiation of microbes by selective growth media, utilizing small molecules (mostly sugars and sugar alcohols), was historically a mainstay of clinical microbiology⁴¹. We (and others) have exploited this differential metabolism by mammalian cells versus pathogens⁴², to develop several pathogen-specific imaging approaches^15,16. While we had originally developed ¹⁸F-FDS, a radioanalog of sorbitol (sugar alcohol), to selectively image infections due to the Enterobacterales group of Gram-negative bacteria^32,33, a study by Kim et al. described the utility of ¹⁸F-FDS PET to detect and monitor invasive pulmonary aspergillosis in animal models²⁷. However, the Kim et al. study could not explain why Aspergillus infections could be visualized well in vivo using ¹⁸F-FDS PET, even though the in vitro uptake of ¹⁸F-FDS is lower in molds compared to E. coli for the same mass. Although growth dynamics and the different morphotypes noted for fungi (but not for bacteria) could affect in vitro uptake, we hypothesized that differences in the biomass occupied by molds versus bacteria within infectious lesions may explain the excellent PET signal noted in vivo. Our study confirmed this, and we demonstrate that the biomass occupied by Aspergillus spp. in pulmonary, cerebral, or rhinosinusal infections is ~50-fold higher than for bacterial infections (Fig. 6). Similar findings were observed in tissues obtained from animal infections. These mechanistic findings have important implications for the development of novel pathogen-specific imaging tracers. Further, Lai et al. have hypothesized that the low uptake of ¹⁸F-FDS by Aspergillus is likely due to the need for induction of sorbitol dehydrogenase by sorbitol, which may not be achieved by PET tracers⁴³. Our study disproves this hypothesis, demonstrating that Aspergillus rapidly accumulates ¹⁸F-FDS without the need for induction by sorbitol, via a saturable process (Fig. 5). Using radio-TLC studies, we further demonstrate that ¹⁸F-FDS is indeed rapidly metabolized intracellularly by A. fumigatus, which is identical to the intracellular metabolism of ¹⁸F-FDS noted in E. coli (Fig. S8).

Our data show that ¹⁸F-FDS-PET can distinguish invasive mold infections from normal tissue and sterile inflammation or cancerous lesions in humans. This is due to low tissue background signal from ¹⁸F-FDS in several tissues, including the lungs and the brain, which are the two primary sites for mold infections. Unlike molds and some other pathogens, mammalian cells cannot metabolize sorbitol, and substitution of the hydroxyl group by fluorine at the second position in ¹⁸F-FDS completely abrogates recognition by mammalian cells⁴⁴. We have previously shown that cancer cell lines do not take up ¹⁸F-FDS³², and that ¹⁸F-FDS PET demonstrates low uptake in cancerous lesions in animal and human studies^32,33. Some brain tumors or inflammatory pathologies have been reported to accumulate ¹⁸F-FDS^45,46,47, but these reports describe pituitary lesions (outside the blood-brain barrier), and / or measure ¹⁸F-FDS uptake immediately after tracer injection (5 min in the case reports^45,46 and 60 min in animal studies⁴⁷), and thus have a higher background signal. Utilizing dynamic ¹⁸F-FDS PET over 120 min, we have demonstrated that while brain tumors can demonstrate initial PET uptake, this signal dissipates over time (after 60 min)³². Therefore, the findings previously noted in brain tumors or inflammatory pathologies are consistent with a non-specific, blood pool effect, consistent with capillary leak at the site of inflammation or tumor⁴⁸. In our studies, imaging was performed 120 or 180 min after tracer injection with significant clearance of the background^32,33,49, with the TNT measurements for invasive mold infections highest in the brain. For example, in patient 4, where imaging was performed 180 min after tracer injection, the ¹⁸F-FDS PET derived TNT in the right frontal cortex lesion was 17 (Fig. S4). The corresponding head MRI (performed for clinical reasons) reported postoperative changes from prior surgeries and white matter changes, but did not demonstrate evidence of pathologic enhancement.

¹⁸F-FDS PET also shows promise for the diagnosis and localization of non-Aspergillus molds, which may have a similar clinical presentation, are on the rise due to the use of Aspergillus-targeted antifungal prophylaxis⁷, and are not detected by current noninvasive biomarkers (i.e., galactomannan)⁶. In addition to Aspergillus spp., we demonstrate rapid ¹⁸F-FDS accumulation in a diverse group of clinical isolates, including azole-resistant molds, confirming that sorbitol uptake is a conserved process. Although, Aspergillus is known to metabolize sorbitol^34,35, ¹⁸F-FDS accumulation by a range of clinically relevant mold fungi is a new finding from our studies. Our findings are further supported by clinically relevant animal models of invasive pulmonary, rhinosinusal and cerebral mold infections, and human studies. In one exemplary case, ¹⁸F-FDS PET was able to localize known pulmonary invasive mold infection due to C. bantiana in a patient with a negative galactomannan test. Here, ¹⁸F-FDS PET also identified a previously undiagnosed cerebral involvement with C. bantiana that was not detected on brain MRI performed for clinical reasons (Fig. S4). C. bantiana has a high propensity for dissemination to the brain and is associated with high mortality due to delays in diagnosis (median of 115 days)⁵⁰. Early and accurate identification of cerebral infection has implications for treatment (the need for an antifungal agent with CNS penetration) and for prognostication. This example highlights many potential advantages of ¹⁸F-FDS PET, including its utility in diagnosing and localizing infections due to diverse clinically important molds not otherwise detected by currently available biomarkers, and a level of sensitivity superior to the best currently available imaging tools. Although larger clinical studies are needed to validate sensitivity, our data suggest that ¹⁸F-FDS PET is expected to have high sensitivity. Due to a lack of other noninvasive diagnostic tools for deep-seated invasive mold infections, the high sensitivity of ¹⁸F-FDS PET, could also be beneficial in ruling out mold infections, which may reduce the use of unnecessary empiric antifungal treatments. To our knowledge, in vivo detection of non-Aspergillus mold infections as well as the use of rhinosinusal and cerebral models of invasive mold disease, including non-Aspergillus molds, are novel and have not been reported previously.

¹⁸F-FDS PET is a promising diagnostic tool that can be accessible to a range of institutions and patient populations. PET is now a routine clinical tool⁵¹, and is increasingly available both in high- and low-income and middle-income countries. Due to the short half-life of most PET radionuclides, e.g., 109 min for F-18, the tracer has to be synthesized on-site or at a nearby location, requiring dedicated infrastructure (cyclotron) and trained personnel, thus limiting on-demand supply. However, ¹⁸F-FDS can be synthesized rapidly from commercially available ¹⁸F-FDG by a simple, one-step kit-based method³⁷, allowing on-demand synthesis (at room temperature) and global availability of ¹⁸F-FDS. Further, high specific activity (radioactivity per mass) is not required, as even low specific activity ¹⁸F-FDS (20–50 Ci/mM) produces optimal imaging characteristics^32,42. We believe this is a major advantage that will allow global, on-demand ¹⁸F-FDS availability at any site with access to a PET scanner. Most patients suspected of invasive mold infections will undergo imaging studies such as CT or MRI as the current standard of care. ¹⁸F-FDS PET/CT could be incorporated into the current clinical workflow for these patients, especially in resource-rich settings. Given the rapid clearance of ¹⁸F-FDS, the radiation exposure from a whole-body ¹⁸F-FDS PET combined with a low-dose CT scan is estimated to be ~1 rem for an adult patient, which is less than the radiation exposures from a contrast-enhanced chest-abdomen-pelvis CT (~2 rem), routinely performed clinically^52,53. PET imaging typically takes 15–60 min, performed 60–120 min after tracer administration, and provides results on the same day. In fact, the ¹⁸F-FDS PET/CT results in the current study were available much earlier than the conventional tests used to establish the confirmatory diagnosis for invasive mold infections. Newer technologies such as total-body PET scanners enable high-sensitivity imaging with dramatically improved signal-to-noise ratio and will further shorten the scan duration while increasing sensitivity⁵⁴.

Our studies have limitations. Many PCR based tests (and fungal cultures) provide species or genus level identification and can also detect mixed mold infections (reported rarely in immunocompromised patients⁵⁵). Although ¹⁸F-FDS PET cannot distinguish aspergillosis from other mold infections (e.g. mucormycosis) an approach combining ¹⁸F-FDS PET with galactomannan testing could aid in distinguishing between the two infections. Further, due to the need for invasive procedures, many mucormycosis cases remain undiagnosed or the diagnoses are significantly delayed. Therefore, ¹⁸F-FDS PET could not only help in the early diagnosis, but when positive, also justify the need for an earlier invasive procedure as well as highlight the best sites(s) to sample the infection, due to its high sensitivity and specificity. Moreover, our prior studies have shown that there is no correlation between the peripheral white blood cell or neutrophil counts with the ¹⁸F-FDS PET signal in the infected patients³³, suggesting that ¹⁸F-FDS PET is not dependent on host inflammatory cells. Additionally, all animal experiments in the current studies utilized neutropenic mice (induced with cyclophosphamide), which did not preclude ¹⁸F-FDS PET to specifically detect, localize and monitor invasive fungal infections. We utilized stringent criteria in this clinical study (i.e., prospective enrollment, eventual confirmation of invasive mold infection from the infection site, and inclusion of patients who received ≤7 days of antifungal treatment prior to the ¹⁸F-FDS PET) strengthening the reliability and clinical relevance of our results, given that empiric antifungal treatments are often initiated before imaging studies or definitive testing for invasive mold infections are performed¹⁴. However, these criteria were also an impediment to patient recruitment, leading to a small sample size for our clinical study. A validated PET tracer in clinical use would not be subject to such study restrictions, so we believe that ¹⁸F-FDS PET will ultimately have applications in many more patients for whom early diagnosis and localization can significantly alter the diagnostic pathway, treatment, and prognosis. Moreover, due to the ease of tracer synthesis and on-demand availability with the one-step, solid-phase cartridge system³⁷, this technology could be practically implemented in most settings. Notably, Enterobacterales bacteria (e.g., E. coli, Salmonella, etc.) also metabolize sorbitol and acute [high burden, 7 log₁₀ colony forming units (CFU)/mL] infections would be visualized by ¹⁸F-FDS PET³³. However, Enterobacterales bacteria are rarely in the differential for presentations consistent with invasive mold infections and patient characteristics, response to empiric antibiotics can easily distinguish Enterobacterales from mold infections. Similarly, while Pseudomonas pneumonia can present in this patient population, Pseudomonas do not accumulate ¹⁸F-FDS, and no ¹⁸F-FDS PET signal was noted in the mouse model of Pseudomonas pneumonia. Future studies will also evaluate whether ¹⁸F-FDS PET can be used to detect other invasive fungal infections, such as Candida and Cryptococcus.

In summary, we present data from a prospective clinical study evaluating whole-body ¹⁸F-FDS PET/CT in nine patients with high-suspicion of invasive mold infections or other pathologies. We demonstrate that ¹⁸F-FDS PET is feasible and can rapidly detect and localize invasive pulmonary and cerebral infections due to Aspergillus, non-Aspergillus, and azole-resistant molds, and differentiate them from sterile inflammation or cancer. Imaging data were confirmed by culture or molecular assays, and observations in humans were replicated using clinically relevant animal models. Animal studies also demonstrated that ¹⁸F-FDS PET can monitor the response to antifungal treatments. While larger clinical studies are needed to validate these findings, our data strongly support the role of ¹⁸F-FDS as an easily synthesizable PET tracer for noninvasive detection and localization of invasive mold infections throughout the body.

Methods

All protocols were approved by the Johns Hopkins University Biosafety, Radiation Safety, Animal Care and Use (MO22M354) and Institutional Review Board (IRB00097331, IRB00071002) Committees.

¹⁸F-FDS PET/CT in prospectively enrolled patients

¹⁸F-FDS was synthesized as a sterile solution with high specific activity and high radiochemical purity by the Johns Hopkins PET Center. This prospective observational study was performed in accordance with the U.S. FDA Radioactive Drug Research Committee guidelines⁵⁶, with localization of ¹⁸F-FDS PET signal to the infection sites as the primary outcome measure. This study was registered on clinicaltrials.gov (NCT05611892)⁵⁷. Nine patients (four with newly diagnosed invasive mold infections, and five control patients with other pathologies³³) were prospectively recruited from the Johns Hopkins Hospitals between October 2016 to July 2024 using the following criteria (Table S1). Confirmatory diagnosis of mold infections was established in all four patients, using culture or molecular assays. Sterile inflammatory pathologies were determined clinically and/or by tissue biopsies, and oncologic diagnosis was made via tissue biopsy (Table S2). Written informed consent was obtained from each subject, physical examination performed by a trained physician and screening laboratory tests performed and reviewed by the study principal investigator to confirm eligibility. On the day of imaging, a low-dose CT, selected to minimize the radiation exposure, was performed. An intravenous injection of ¹⁸F-FDS (503 ± 191 MBq) was administered and PET was acquired at 120 min after tracer administration (except patient 4 where acquisition was at 180 min post-tracer injection) from the skull vertex to mid-thigh with 3 min per bed-position, using a Siemens Biograph PET/CT scanner. All patients were evaluated immediately after finishing the PET scan and 20-25 days to assess early and delayed adverse events.

Human histology

Infected tissue samples from eight unique patients with confirmed pulmonary, cerebral or rhinosinusal invasive Aspergillus spp. infections from Johns Hopkins Hospitals were obtained (patient characteristics outlined in Table S3). Multiple random slides were obtained from each tissue sample, stained with GMS, digitally scanned and analyzed at 40x magnification. Additionally, publicly available digitized slides of bacterial infections in patients (case IDs: 963, 749, 1692, 1688) from Infectious Diseases Images eMicrobes Digital Library were analyzed³⁸. The percentage of area occupied by the mold or bacteria was calculated using automated segmentation (ImageJ) after standardizing the area analyzed in each sample (Fig. 6, S10).

In vitro assays

A. fumigatus (ATCC 1022), or clinical isolates from the Johns Hopkins Hospitals clinical microbiology laboratory including Rhizopus arrhizus and Mucor circinelloides were grown on potato dextrose agar (PDA) plates at 37 °C and conidia were collected in sterile phosphate buffered saline (PBS) with Tween 80 (0.01%). A manual count of the conidia density (conidia/mL) was performed using a hemocytometer. Fungal cultures were heat killed by exposure to 95 °C heat for 60 min. ¹⁸F-FDS (5 μCi/mL) was co-incubated with the fungi grown at 37 °C with shaking and samples were collected at different time points followed by pelleting and washing thrice with PBS. Activity for each sample pellet was measured immediately using an automated gamma counter (PerkinElmer). Counts were corrected for background and decay and normalized to pellet weight (mg). Bacterial and mammalian cells (THP-1) studies were performed as described before³². All studies were performed in triplicate samples.

Metabolite studies

A. fumigatus (ATCC 1022) or E. coli (ATCC 25922) cultures were incubated with ¹⁸F-FDS. Aliquots of 1 mL were pelleted by centrifugation and washed with PBS. The cells were lysed and the cytoplasmic fraction was separated by ultracentrifugation and filtered (0.22 µm syringe). The solute was analyzed on a radio-TLC (Eckert & Ziegler) using ¹⁸F-FDS and ¹⁸F-FDG as reference.

Animal studies

Female BALB/c mice (6-8 weeks old, Charles River) were housed in pre-sterilized, filter-topped cages, with unrestricted access to both food and water [medicated to prevent bacterial superinfection prophylaxis with tetracycline (Sigma)]. All animals received anesthesia with 2% isoflurane and O₂ mixture for infection inoculation, LPS-induced inflammation and imaging.

For the mold infection studies, mice were immunosuppressed using cortisone acetate (Sigma, 200 mg/Kg) administered subcutaneously at day five and one prior to mold infection and cyclophosphamide (Sigma, 150 mg/Kg) administered intraperitoneally at day four and one day before infection^27,58. For invasive pulmonary aspergillosis, anesthetized mice were hung by their upper incisors and secured to a support. A superficial 1 cm incision was made in the midline of the neck to visualize and successfully cannulate the trachea⁵⁹. The intubation was confirmed upon connection to a ventilator, and then 40 µL suspension of fungal conidia [7 × 10⁶A. fumigatus (ATCC 1022) conidia/mouse] was used for the infection followed by two days of incubation prior to imaging. Rhinosinual infection was carried out by a slow intranasal instillation of 15 µL of the fungal conidia suspension [1 × 10⁸ A. fumigatus (ATCC 1022), Rhizopus arrhizus or Mucor circinelloides conidia per mouse) under anesthesia with two days of incubation prior to imaging. M. circinelloides is one of the most common mucoralean fungi isolated from clinical samples⁶⁰, and causes human mucormycosis including fungal rhinosinusitis⁶¹. In a subset of mice with rhinosinusal Aspergillus-infection, voriconazole (MedChem; 40 mg/kg/day) was administered daily (5-days a week) via oral gavage for one week, starting two days after infection. For invasive cerebral aspergillosis, anesthetized mice were infected with 3 µL suspension of fungal conidia [1 × 10⁶A. fumigatus (ATCC 1022) conidia/mouse] via a burr hole using a Hamilton syringe (Hamilton, 88,000) and stereotaxic instrument (David KOPF instrument, model 900) with the following coordinates: 2 mm dorsal to bregma, 2 mm lateral to middle line and 2 mm deep. After the inoculation, the hole was sealed with bone wax and the scalp was closed using a surgical staple and the infection incubated for one day prior to imaging. The model of sterile pneumonia was developed using an intratracheal instillation of 30 µL of LPS solution (5 µg/g). For the Pseudomonas pneumonia model, mice were immunosuppressed with cyclophosphamide (Sigma, 150 mg/kg, administered intraperitoneally on day 4 and one day before infection). On the day of infection 40 µL of a Pseudomonas aeruginosa (ATCC 10145) suspension (2 × 10⁵ CFU/mouse) was administered via intratracheal instillation, and imaging performed 3 days later. The pulmonary bacterial burden was assessed immediately after imaging. For systemic inflammation mice were intraperitoneally administered 50 µL of LPS solution four days prior (2 µg/g) and the day of imaging (6 µg/g)^62,63.

Imaging: Animals were injected via tail vain with a bolus of 8.89 MBq ± 1.43 of ¹⁸F-FDG or 8.54 ± 1.56 MBq of ¹⁸F-FDS. A 15 min PET was acquired 45 min after injection of ¹⁸F-FDG, and 120 min after injection of ¹⁸F-FDS, using a nanoScan PET/CT (Mediso). A CT scan was obtained for anatomical reference. MRI was performed using a 7 T preclinical MRI (MR Solutions) and mouse ¹H body coils with 1 mm slices and fat saturation. The T1W and T2W sequences utilized a repetition time of 1000 ms and an echo time of 11 and 45 ms, respectively.

Ex vivo Biodistribution: These studies were performed immediately after the completion of the imaging studies. Animals were sacrificed, and the tissues were perfused by infusing 6 mL of PBS per animal through the left ventricle. The organs were extracted thereafter and the activity in each organ was measured immediately using an automated gamma counter (PerkinElmer) and presented as the percentage injected dose per gram of tissue (%ID/g).

Histopathology: Slides were prepared from tissues fixed in 4% paraformaldehyde from our current mold infection animal models. Multiple random slides were obtained from each tissue sample, stained with GMS, digitally scanned and analyzed as described above. Fixed tissue samples from mice with E. coli myositis with a target implantation of 7 log₁₀ CFU units were obtained from our prior studies³². Multiple random slides were obtained from each tissue sample, Gram stained, digitally scanned and analyzed as described above.

Image analysis

Human images were analyzed using Mirada XD™ 3.6.8 (Mirada Medical) and PMOD version 3.402 (PMOD Technologies LLC), while the animal images were analyzed using VivoQuant 2020 (Invicro)^32,33,64. Three-dimensional VOIs were drawn manually using the CT as a reference and the PET activity was measured. When multiple VOIs were drawn in the same patients, they were from independent lesions or different parts of large lesions (with no overlap of any VOI). The human PET data are presented as TNT which is the ratio of the mean PET signal at the sites of pathology to the unaffected sites of the same tissue. The animal PET data are presented as SUVr which is the ratio of the mean PET signal at the sites of pathology to the same tissue from uninfected animals. All data were visualized using AMIRA 5.2.1 (Visage Imaging, Inc.) or AMIDE 1.0.6 (Andreas Loening).

Statistical analysis

Data were analyzed using Prism 10.0 (GraphPad Software Inc). All data are represented as median ± IQR or mean ± SD and comparisons were made using one-tailed or two-tailed Mann-Whitney U test. P values ≤ 0.05 were considered statistically significant. No sex-based analyses were performed because the study design was not explicitly set up to examine sex differences.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data are available in the main text or the supplementary materials. Source data file has been provided with this paper. Source data are provided with this paper.

References

McCarthy, M., Rosengart, A., Schuetz, A. N., Kontoyiannis, D. P. & Walsh, T. J. Mold infections of the central nervous system. N. Engl. J. Med 371, 150–160 (2014).
Article PubMed PubMed Central Google Scholar
Lee, S. O. Diagnosis and treatment of invasive mold diseases. Infect. Chemother. 55, 10–21 (2023).
Article CAS PubMed Google Scholar
Denning, D.W. Global incidence and mortality of severe fungal disease. Lancet Infect. Dis. 24, e428–e438 (2024).
Article PubMed Google Scholar
Hoenigl, M. et al. COVID-19-associated fungal infections. Nat. Microbiol 7, 1127–1140 (2022).
Article CAS PubMed PubMed Central Google Scholar
Pasquier, G. COVID-19-associated mucormycosis in India: why such an outbreak?. J. Mycol. Med 33, 101393 (2023).
Article PubMed PubMed Central Google Scholar
Lamoth, F. & Kontoyiannis, D. P. Therapeutic challenges of non-aspergillus invasive mold infections in immunosuppressed patients. Antimicrob. Agents Chemother. 63, e01244–e01319 (2019).
Article CAS PubMed PubMed Central Google Scholar
Wattier, R. L. et al. A prospective, International Cohort Study of invasive mold infections in children. J. Pediatr. Infect. Dis. Soc. 4, 313–322 (2015).
Article Google Scholar
Hoenigl, M. et al. Global guideline for the diagnosis and management of rare mould infections: an initiative of the European Confederation of Medical Mycology in cooperation with the International Society for Human and Animal Mycology and the American Society for Microbiology. Lancet Infect. Dis. 21, e246–e257 (2021).
Article CAS PubMed Google Scholar
World Health Organization. WHO fungal priority pathogens list to guide research, development and public health action. https://www.who.int/publications/i/item/9789240060241 (2022).
Bergin, S. P. et al. Plasma microbial cell-free DNA sequencing in immunocompromised patients with pneumonia: a prospective observational study. Clin. Infect. Dis. 78, 775–784 (2024).
Article CAS PubMed Google Scholar
Blauwkamp, T. A. et al. Analytical and clinical validation of a microbial cell-free DNA sequencing test for infectious disease. Nat. Microbiol 4, 663–674 (2019).
Article CAS PubMed Google Scholar
Mercier, T. et al. Defining galactomannan positivity in the updated EORTC/MSGERC consensus definitions of invasive fungal diseases. Clin. Infect. Dis. 72, S89–S94 (2021).
Article ADS CAS PubMed Google Scholar
Zhou, W. et al. Diagnostic value of galactomannan antigen test in serum and bronchoalveolar lavage fluid samples from patients with nonneutropenic invasive pulmonary aspergillosis. J. Clin. Microbiol 55, 2153–2161 (2017).
Article CAS PubMed PubMed Central Google Scholar
Ordonez, A.A., et al. Molecular imaging of bacterial infections: Overcoming the barriers to clinical translation. Sci. Transl. Med. 11, eaax8251 (2019).
Article PubMed PubMed Central Google Scholar
Chen, X. et al. Visualizing bacterial infections with novel targeted molecular imaging approaches. J. Infect. Dis. 228, S249–S258 (2023).
Article PubMed PubMed Central Google Scholar
Lai, J., Wang, B., Petrik, M., Beziere, N. & Hammoud, D. A. Radiotracer development for fungal-specific imaging: past, present, and future. J. Infect. Dis. 228, S259–S269 (2023).
Article PubMed PubMed Central Google Scholar
Kuhlman, J. E. et al. CT of invasive pulmonary aspergillosis. AJR Am. J. Roentgenol. 150, 1015–1020 (1988).
Article CAS PubMed Google Scholar
Leroy-Freschini, B. et al. 18F-FDG PET/CT for invasive fungal infection in immunocompromised patients. QJM 111, 613–622 (2018).
Article CAS PubMed Google Scholar
Ankrah, A.O., et al. The Added Value of [(18)F]FDG PET/CT in the Management of Invasive Fungal Infections. Diagnostics 11, 137 (2021).
Article CAS PubMed PubMed Central Google Scholar
Sharma, P., Mukherjee, A., Karunanithi, S., Bal, C. & Kumar, R. Potential role of 18F-FDG PET/CT in patients with fungal infections. AJR Am. J. Roentgenol. 203, 180–189 (2014).
Article PubMed Google Scholar
Petrik, M. et al. 68Ga-siderophores for PET imaging of invasive pulmonary aspergillosis: proof of principle. J. Nucl. Med. 51, 639–645 (2010).
Article CAS PubMed Google Scholar
Petrik, M. et al. 68Ga-triacetylfusarinine C and 68Ga-ferrioxamine E for Aspergillus infection imaging: uptake specificity in various microorganisms. Mol. Imaging Biol. 16, 102–108 (2014).
Article PubMed Google Scholar
Rolle, A. M. et al. ImmunoPET/MR imaging allows specific detection of Aspergillus fumigatus lung infection in vivo. Proc. Natl Acad. Sci. USA 113, E1026–1033 (2016).
Article CAS PubMed PubMed Central Google Scholar
Davies, G. et al. Towards translational immunoPET/MR imaging of invasive pulmonary aspergillosis: the humanised monoclonal antibody JF5 detects aspergillus lung infections in vivo. Theranostics 7, 3398–3414 (2017).
Article CAS PubMed PubMed Central Google Scholar
Thornton, C. R. Molecular imaging of invasive pulmonary aspergillosis using immunoPET/MRI: the future looks bright. Front. Microbiol. 9, 691 (2018).
Article PubMed PubMed Central Google Scholar
Henneberg, S., et al. Antibody-guided in vivo imaging of Aspergillus fumigatus lung infections during antifungal azole treatment. Nat. Commun. 12, 1707 (2021).
Article ADS PubMed PubMed Central Google Scholar
Kim, D. Y., et al. In vivo imaging of invasive aspergillosis with (18)F-fluorodeoxysorbitol positron emission tomography. Nat. Commun. 13, 1926 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Lai, J. et al. PET imaging of Aspergillus infection using Zirconium-89 labeled anti-beta-glucan antibody fragments. Eur. J. Nucl. Med Mol. Imaging 51, 3223–3234 (2024).
Article CAS PubMed PubMed Central Google Scholar
Shah, S. et al. Development and preclinical validation of 2-deoxy 2-[(18)F]fluorocellobiose as an Aspergillus-specific PET tracer. Sci. Transl. Med. 16, eadl5934 (2024).
Article CAS PubMed Google Scholar
Basuli, F. et al. Fully automated cassette-based synthesis of 2-Deoxy-2-[(18)F]fluorocellobiose using trasis allinone module. J. Label. Comp. Radiopharm. 67, 308–313 (2024).
Article CAS Google Scholar
Schwenck, J., et al. Antibody-guided molecular imaging of aspergillus lung infections in leukemia patients. J. Nucl. Med. https://doi.org/10.2967/jnumed.121.263251 (2022).
Weinstein, E. A. et al. Imaging Enterobacteriaceae infection in vivo with 18F-fluorodeoxysorbitol positron emission tomography. Sci. Transl. Med 6, 259ra146 (2014).
Article PubMed PubMed Central Google Scholar
Ordonez, A.A., et al. Imaging Enterobacterales infections in patients using pathogen-specific positron emission tomography. Sci. Transl. Med. 13, eabe9805 (2021).
Article CAS PubMed PubMed Central Google Scholar
Elshafei, A. M. & Abdel-Fatah, O. M. Evidence for a non-phosphorylated route of galactose breakdown in cell-free extracts of Aspergillus niger. Enzym. Micro. Technol. 29, 76–83 (2001).
Article CAS Google Scholar
Koivistoinen, O. M., Richard, P., Penttila, M., Ruohonen, L. & Mojzita, D. Sorbitol dehydrogenase of Aspergillus niger, SdhA, is part of the oxido-reductive D-galactose pathway and essential for D-sorbitol catabolism. FEBS Lett. 586, 378–383 (2012).
Article CAS PubMed Google Scholar
Li, Z. B. et al. The synthesis of 18F-FDS and its potential application in molecular imaging. Mol. Imaging Biol. 10, 92–98 (2008).
Article PubMed Google Scholar
Mota, F., De Jesus, P. & Jain, S. K. Kit-based synthesis of 2-deoxy-2-[(18)F]-fluoro-D-sorbitol for bacterial imaging. Nat. Protoc. 16, 5274–5286 (2021).
Article CAS PubMed PubMed Central Google Scholar
Partners’ Infectious Disease Images. Infectious Diseases Images eMicrobes Digital Library. https://www.idimages.org/atlas (2024).
DeLone, D. R. et al. Disseminated aspergillosis involving the brain: distribution and imaging characteristics. AJNR Am. J. Neuroradiol. 20, 1597–1604 (1999).
CAS PubMed PubMed Central Google Scholar
Tempkin, A. D., Sobonya, R. E., Seeger, J. F. & Oh, E. S. Cerebral aspergillosis: radiologic and pathologic findings. Radiographics 26, 1239–1242 (2006).
Article PubMed Google Scholar
Mac Faddin, J.F. Biochemical tests for identification of medical bacteria, (Williams & Wilkins, Baltimore, 1976).
Ordonez, A. A. et al. A systematic approach for developing bacteria-specific imaging tracers. J. Nucl. Med. 58, 144–150 (2017).
Article CAS PubMed PubMed Central Google Scholar
Lai, J. et al. Evaluation of 2-[(18)F]-fluorodeoxysorbitol PET imaging in preclinical models of aspergillus infection. J. Fungi 8, 25 (2021).
Article Google Scholar
Scott, M. E. & Viola, R. E. The use of fluoro- and deoxy-substrate analogs to examine binding specificity and catalysis in the enzymes of the sorbitol pathway. Carbohydr. Res. 313, 247–253 (1998).
Article CAS PubMed Google Scholar
Kong, Z., Wang, Y. & Ma, W. & Cheng, X. Role of (18)F-fluorodeoxyglucose (FDG) and (18)F-2-fluorodeoxy sorbitol (FDS) in autoimmune hypophysitis: a case report. BMC Endocr. Disord. 20, 84 (2020).
Article PubMed PubMed Central Google Scholar
Cheng, X., Zhu, W. & Cui, R. Increased 18F-2-fluorodeoxysorbitol (18F-FDS) activity in a pituitary spindle cell carcinoma. Clin. Nucl. Med 41, 953–955 (2016).
Article PubMed Google Scholar
Hugon, G., et al. [(18)F]2-fluoro-2-deoxy-sorbitol PET imaging for quantitative monitoring of enhanced blood-brain barrier permeability induced by focused ultrasound. Pharmaceutics 13, 1752 (2021).
Article CAS PubMed PubMed Central Google Scholar
Javadi, M. S., Ordonez, A. A. & Jain, S. K. Re: Increased 18F-2-fluorodeoxysorbitol (18F-FDS) activity in a pituitary spindle cell carcinoma. Clin. Nucl. Med 42, 649 (2017).
Article PubMed PubMed Central Google Scholar
Zhu, W., et al. Biodistribution and Radiation Dosimetry of the Enterobacteriaceae-Specific Imaging Probe [F]Fluorodeoxysorbitol Determined by PET/CT in Healthy Human Volunteers. Mol. Imaging Biol. 18, 782–787 (2016).
Article PubMed Google Scholar
Chakrabarti, A. et al. Brain abscess due to Cladophialophora bantiana: a review of 124 cases. Med Mycol. 54, 111–119 (2016).
Article PubMed Google Scholar
Hricak, H. et al. Medical imaging and nuclear medicine: a Lancet Oncology Commission. Lancet Oncol. 22, e136–e172 (2021).
Article PubMed PubMed Central Google Scholar
Wright, W. F. & Auwaerter, P. G. Fever and fever of unknown origin: review, recent advances, and lingering dogma. Open Forum Infect. Dis. 7, ofaa132 (2020).
Article PubMed PubMed Central Google Scholar
Bharucha, T. et al. Diagnostic yield of FDG-PET/CT in fever of unknown origin: a systematic review, meta-analysis, and Delphi exercise. Clin. Radio. 72, 764–771 (2017).
Article CAS Google Scholar
Henrich, T. J., Jones, T., Beckford-Vera, D., Price, P. M. & VanBrocklin, H. F. Total-body PET imaging in infectious diseases. PET Clin. 16, 89–97 (2021).
Article PubMed Google Scholar
McLintock, L. A., Gibson, B. E. & Jones, B. L. Mixed pulmonary fungal infection with Aspergillus fumigatus and Absidia corymbifera in a patient with relapsed acute myeloid leukaemia. Br. J. Haematol. 128, 737 (2005).
Article PubMed Google Scholar
FDA. Guidance for industry and researchers: the radioactive drug research committee: human research without an investigational new drug application. (2010).
University, J.H. Biodistribution and pathophysiology study of 18F-fluorodeoxysorbitol (18F-FDS) in diseased patients. (ClinicalTrials.gov, 2024).
Wang, Y. et al. Detection of Aspergillus fumigatus pulmonary fungal infections in mice with (99m)Tc-labeled MORF oligomers targeting ribosomal RNA. Nucl. Med Biol. 40, 89–96 (2013).
Article PubMed Google Scholar
Villabona-Rueda, A., Wang, D. & D’Alessio, F.R. Experimental model to evaluate resolution of pneumonia. J. Vis. Exp. https://doi.org/10.3791/63925 (2023).
Badali, H. et al. Epidemiology and antifungal susceptibilities of mucoralean fungi in clinical samples from the United States. J. Clin. Microbiol 59, e0123021 (2021).
Article PubMed Google Scholar
Erami, M., et al. Microbial and clinical epidemiology of invasive fungal rhinosinusitis in hospitalized COVID-19 patients, the divergent causative agents. Med. Mycol. 61, myad020 (2023).
Article PubMed Google Scholar
Liverani, E. et al. LPS-induced systemic inflammation is more severe in P2Y12 null mice. J. Leukoc. Biol. 95, 313–323 (2014).
Article PubMed PubMed Central Google Scholar
Xingi, E., et al. LPS-induced systemic inflammation affects the dynamic interactions of astrocytes and microglia with the vasculature of the mouse brain cortex. Cells 12, 1418 (2023).
Article CAS PubMed PubMed Central Google Scholar
Chen, X., et al. Dynamic PET reveals compartmentalized brain and lung tissue antibiotic exposures of tuberculosis drugs. Nat. Commun. 15, 6657 (2024).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgments

We thank all the subjects who participated in the study. We would also like to thank Jeff Leal (Johns Hopkins Hospitals) for curating the human imaging data, Kelly Flavahan (Johns Hopkins Hospitals) for assistance with PET/CT imaging, and Warda Memon (Johns Hopkins Hospitals) for assistance with curating the clinical isolates. This work was funded by the U.S. National Institutes of Health R01-AI145435-A1 (S.K.J.), R01-AI153349 (S.K.J.), R01-HL131829 (S.K.J.), S10-OD030381-A1 (S.K.J.), S10-OD034232 (M.A.L.), K08-AI139371 (E.W.T.) and funds from the Hankins Young Investigator Award (S.K.J.).

Author information

These authors contributed equally: Carlos E. Ruiz-Gonzalez, Oscar J. Nino-Meza, Medha Singh.

Authors and Affiliations

Center for Infection and Inflammation Imaging Research, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Carlos E. Ruiz-Gonzalez, Oscar J. Nino-Meza, Medha Singh, Yuderleys Masias-Leon, Amy Kronenberg, Madelynn Shambles, Xueyi Chen, Elizabeth W. Tucker & Sanjay K. Jain
Department of Pediatrics, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Carlos E. Ruiz-Gonzalez, Oscar J. Nino-Meza, Medha Singh, Yuderleys Masias-Leon, Xueyi Chen & Sanjay K. Jain
Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Elizabeth W. Tucker
Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Martin A. Lodge, Laurence S. Carroll & Sanjay K. Jain
Department of Oncology, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Kenneth R. Cooke
Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Olivia S. Kates & Shmuel Shoham
Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Sean X. Zhang

Authors

Carlos E. Ruiz-Gonzalez
View author publications
Search author on:PubMed Google Scholar
Oscar J. Nino-Meza
View author publications
Search author on:PubMed Google Scholar
Medha Singh
View author publications
Search author on:PubMed Google Scholar
Yuderleys Masias-Leon
View author publications
Search author on:PubMed Google Scholar
Amy Kronenberg
View author publications
Search author on:PubMed Google Scholar
Madelynn Shambles
View author publications
Search author on:PubMed Google Scholar
Xueyi Chen
View author publications
Search author on:PubMed Google Scholar
Elizabeth W. Tucker
View author publications
Search author on:PubMed Google Scholar
Martin A. Lodge
View author publications
Search author on:PubMed Google Scholar
Laurence S. Carroll
View author publications
Search author on:PubMed Google Scholar
Kenneth R. Cooke
View author publications
Search author on:PubMed Google Scholar
Olivia S. Kates
View author publications
Search author on:PubMed Google Scholar
Shmuel Shoham
View author publications
Search author on:PubMed Google Scholar
Sean X. Zhang
View author publications
Search author on:PubMed Google Scholar
Sanjay K. Jain
View author publications
Search author on:PubMed Google Scholar

Contributions

C.E.R-G. and S.K.J. conceptualized and designed the studies. M.S., A.K., and ML.S. performed the in vitro studies. M.S. and L.S.C. performed bacterial metabolite studies. C.E.R-G. developed the animal models and performed the animal studies with help from Y.M-L. and X.C. O.J.N-M., E.W.T., M.A.L., and S.K.J. wrote the protocol for the ¹⁸F-FDS PET clinical studies with support from S.S. and S.X.Z. O.J.N-M. and X.C. coordinated the clinical studies and E.W.T., K.R.C., O.S.K., and S.K.J. recruited and consented the patients. S.X.Z. provided clinical isolates and human histology slides. C.E.R-G. analyzed the data with support from O.J.N-M., M.S., and Y.M-L. S.K.J. provided funding and supervised the project. C.E.R-G. and S.K.J. wrote the manuscript with substantial input from all co-authors.

Corresponding authors

Correspondence to Sean X. Zhang or Sanjay K. Jain.

Ethics declarations

Competing interests

S.K.J. is a co-inventor on pending patent USPA #63/071,755 on the solid-phase cartridge to formulate ready-to-use ¹⁸F-FDS and pending patent USPA #63/699,572 on Noninvasive Detection of Invasive Mold Infections by ¹⁸F-FDS PET, filed by Johns Hopkins University. All other authors declare that they have no competing interests.

Peer review

Peer review information

Nature Communications thanks Jung-Joon Min, and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Reporting summary (download PDF )

Transparent Peer Review file (download PDF )

Source data

Source Data (download XLSX )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Ruiz-Gonzalez, C.E., Nino-Meza, O.J., Singh, M. et al. ¹⁸F-Fluorodeoxysorbitol PET for noninvasive detection of invasive mold infections: preclinical and first-in-human studies. Nat Commun 16, 6395 (2025). https://doi.org/10.1038/s41467-025-61700-6

Download citation

Received: 28 March 2025
Accepted: 30 June 2025
Published: 10 July 2025
Version of record: 10 July 2025
DOI: https://doi.org/10.1038/s41467-025-61700-6